Abstract
We develop a general mathematical framework for variational problems where the unknown function takes values in the space of probability measures on some metric space. We study weak and strong topologies and define a total variation seminorm for functions taking values in a Banach space. The seminorm penalizes jumps and is rotationally invariant under certain conditions. We prove existence of a minimizer for a class of variational problems based on this formulation of total variation and provide an example where uniqueness fails to hold. Employing the Kantorovich–Rubinstein transport norm from the theory of optimal transport, we propose a variational approach for the restoration of orientation distribution function-valued images, as commonly used in diffusion MRI. We demonstrate that the approach is numerically feasible on several data sets.
Similar content being viewed by others
Notes
The normed space \((\mathcal {M}_0(X), \Vert \cdot \Vert _{K\!R})\) is not complete unless X is a finite set [79, Proposition 2.3.2]. Instead, the completion of \((\mathcal {M}_0(X), \Vert \cdot \Vert _{K\!R})\) that we denote here by \(K\!R(X)\) is isometrically isomorphic to the Arens–Eells space AE(X).
References
Aganj, I., Lenglet, C., Sapiro, G.: ODF reconstruction in Q-ball imaging with solid angle consideration. In: Proceedings of the IEEE International Symposium on Biomed Imaging 2009, pp. 1398–1401 (2009)
Ahrens, C., Nealy, J., Pérez, F., van der Walt, S.: Sparse reproducing kernels for modeling fiber crossings in diffusion weighted imaging. In: Proceedings of the IEEE International Symposium on Biomed Imaging 2013, pp. 688–691 (2013)
Ambrosio, L.: Metric space valued functions of bounded variation. Ann. Sc. Norm. Super. Pisa Cl. Sci. IV. Ser. 17(3), 439–478 (1990)
Ambrosio, L., Fusco, N., Pallara, D.: Functions of Bounded Variation and Free Discontinuity Problems. Clarendon Press, Oxford (2000)
Åström, F., Petra, S., Schmitzer, B., Schnörr, C.: Image labeling by assignment. J. Math. Imaging Vis. 58(2), 211–238 (2017)
Ball, J.: A version of the fundamental theorem for Young measures. In: PDEs and Continuum Models of Phase Transitions. Proceedings of an NSF-CNRS Joint Seminar Held in Nice, France, January 18–22, 1988, pp. 207–215 (1989)
Basser, P.J., Mattiello, J., LeBihan, D.: MR diffusion tensor spectroscopy and imaging. Biophys. J. 66(1), 259–267 (1994)
Bačák, M., Bergmann, R., Steidl, G., Weinmann, A.: A second order non-smooth variational model for restoring manifold-valued images. SIAM J. Sci. Comput. 38(1), A567–A597 (2016)
Becker, S., Tabelow, K., Voss, H.U., Anwander, A., Heidemann, R.M., Polzehl, J.: Position-orientation adaptive smoothing of diffusion weighted magnetic resonance data (POAS). Med. Image Anal. 16(6), 1142–1155 (2012)
Bourbaki, N.: Integration. Springer, Berlin (2004)
Callaghan, P.T.: Principles of Nuclear Magnetic Resonance Microscopy. Clarendon Press, Oxford (1991)
Canales-Rodríguez, E.J., Daducci, A., Sotiropoulos, S.N., Caruyer, E., Aja-Fernández, S., Radua, J., et al.: Spherical deconvolution of multichannel diffusion MRI data with non-Gaussian noise models and spatial regularization. PLoS ONE 10(10), 1–29 (2015)
Carothers, N.L.: Real Analysis. Cambridge University Press, Cambridge (2000)
Chambolle, A., Caselles, V., Cremers, D., Novaga, M., Pock, T.: An introduction to total variation for image analysis. Theor. Found. Numer. Methods Sparse Recovery 9, 263–340 (2010)
Chambolle, A., Pock, T.: A first-order primal-dual algorithm for convex problems with applications to imaging. J. Math. Imaging Vis. 40(1), 120–145 (2011)
Chambolle, A., Pock, T.: Total roto-translational variation. Technical Report arXiv:1709.09953, arXiv (2017)
Chan, T.F., Esedoglu, S.: Aspects of total variation regularized \(L^1\) function approximation. SIAM J. Appl. Math. 65(5), 1817–1837 (2005)
Chen, D., Mirebeau, J.M., Cohen, L.D.: Global minimum for a finsler elastica minimal path approach. Int. J. Comput. Vis. 122(3), 458–483 (2016). https://doi.org/10.1007/s11263-016-0975-5
Clarke, F.: Functional Analysis, Calculus of Variations and Optimal Control. Springer, London (2013)
Creusen, E., Duits, R., Vilanova, A., Florack, L.: Numerical schemes for linear and non-linear enhancement of DW-MRI. Numer. Math. Theor. Methods Appl. 6(1), 138–168 (2013)
Cuturi, M.: Sinkhorn distances: Lightspeed computation of optimal transport. In: Burges, C.J.C., Bottou, L., Welling, M., Ghahramani, Z., Weinberger, K.Q. (eds.) Advances in Neural Information Processing Systems 26, pp. 2292–2300. Curran Associates, Inc. (2013)
Dacorogna, B.: Direct Methods in the Calculus of Variations, 2nd edn. Springer, Berlin (2008)
Daducci, A., et al.: Quantitative comparison of reconstruction methods for intra-voxel fiber recovery from diffusion MRI. IEEE Trans. Med. Imaging 33(2), 384–399 (2014)
Daducci, A., Canales-Rodríguez, E.J., Descoteaux, M., Garyfallidis, E., Gur, Y., et al.: Quantitative comparison of reconstruction methods for intra-voxel fiber recovery from diffusion MRI. IEEE Trans. Med. Imaging 33(2), 384–399 (2014)
Delputte, S., Dierckx, H., Fieremans, E., D’Asseler, Y., Achten, R., Lemahieu, I.: Postprocessing of brain white matter fiber orientation distribution functions. In: Proceedings of the IEEE International Symposium on Biomed Imaging 2007, pp. 784–787 (2007)
Descoteaux, M.: High angular resolution diffusion MRI: from local estimation to segmentation and tractography. Ph.D. thesis, University of Nice-Sophia Antipolis (2008)
Duchoň, M., Debiève, C.: Functions with bounded variation in locally convex space. Tatra Mt. Math. Publ. 49, 89–98 (2011)
Duits, R., Franken, E.: Left-invariant diffusions on the space of positions and orientations and their application to crossing-preserving smoothing of HARDI images. Int. J. Comput. Vis. 92(3), 231–264 (2011)
Duits, R., Haije, T.D., Creusen, E., Ghosh, A.: Morphological and linear scale spaces for fiber enhancement in DW-MRI. J. Math. Imaging Vis. 46(3), 326–368 (2012)
Duval, V., Aujol, J.F., Gousseau, Y.: The TVL1 model: a geometric point of view. Multiscale Model. Simul. 8(1), 154–189 (2009)
Ehricke, H.H., Otto, K.M., Klose, U.: Regularization of bending and crossing white matter fibers in MRI Q-ball fields. Magn. Reson. Imaging 29(7), 916–926 (2011)
Fitschen, J.H., Laus, F., Schmitzer, B.: Optimal transport for manifold-valued images. In: 2017 Scale Space and Variational Methods in Computer Vision, pp. 460–472 (2017)
Fitschen, J.H., Laus, F., Steidl, G.: Transport between RGB images motivated by dynamic optimal transport. J. Math. Imaging Vis. 56(3), 409–429 (2016)
Garyfallidis, E., Brett, M., Amirbekian, B., Rokem, A., Van Der Walt, S., Descoteaux, M., Nimmo-Smith, I., Contributors, D.: Dipy, a library for the analysis of diffusion MRI data. Front. Neuroinform. 8(8), 1–17 (2014)
Goh, A., Lenglet, C., Thompson, P.M., Vidal, R.: Estimating orientation distribution functions with probability density constraints and spatial regularity. In: Medical Image Computing and Computer-Assisted Intervention—MICCAI 2009, pp. 877–885 (2009)
Goldluecke, B., Strekalovskiy, E., Cremers, D.: The natural vectorial total variation which arises from geometric measure theory. SIAM J. Imaging Sci. 5(2), 537–563 (2012)
Goldstein, T., Esser, E., Baraniuk, R.: Adaptive primal dual optimization for image processing and learning. In: Proceedings of the 6th NIPS Workshop on Optimization for Machine Learning (2013)
Goldstein, T., Li, M., Yuan, X.: Adaptive primal-dual splitting methods for statistical learning and image processing. In: Cortes, C., Lawrence, N.D., Lee, D.D., Sugiyama, M., Garnett, R. (eds.) Advances in Neural Information Processing Systems, vol. 28, pp. 2089–2097. Curran Associates, Inc., New York (2015)
Goldstein, T., Li, M., Yuan, X., Esser, E., Baraniuk, R.: Adaptive primal-dual hybrid gradient methods for saddle-point problems. Technical Report arXiv:1305.0546v2, arXiv (2015)
Hewitt, E., Stromberg, K.: Real and Abstract Analysis. Springer, Berlin (1965)
Hohage, T., Rügge, C.: A coherence enhancing penalty for diffusion MRI: regularizing property and discrete approximation. SIAM J. Imaging Sci. 8(3), 1874–1893 (2015)
Tulcea, A.I., Tulcea, C.I.: Topics in the Theory of Lifting. Springer, Berlin (1969)
Kaden, E., Kruggel, F.: A reproducing kernel hilbert space approach for Q-ball imaging. IEEE Trans. Med. Imaging 30(11), 1877–1886 (2011)
Kantorovich, L.V., Rubinshtein, G.S.: On a functional space and certain extremum problems. Dokl. Akad. Nauk SSSR 115, 1058–1061 (1957)
Karayumak, S.C., Özarslan, E., Unal, G.: Asymmetric orientation distribution functions (AODFs) revealing intravoxel geometry in diffusion MRI. Magn. Reson. Imaging 49, 145–158 (2018)
Kezele, I., Descoteaux, M., Poupon, C., Abrial, P., Poupon, F., Mangin, J.F.: Multiresolution decomposition of HARDI and ODF profiles using spherical wavelets. In: Presented at the Workshop on Computational Diffusion MRI, MICCAI, New York, pp. 225–234 (2008)
Kim, Y., Thompson, P.M., Vese, L.A.: HARDI data denoising using vectorial total variation and logarithmic barrier. Inverse Probl. Imaging 4(2), 273–310 (2010)
Laude, E., Möllenhoff, T., Moeller, M., Lellmann, J., Cremers, D.: Sublabel-accurate convex relaxation of vectorial multilabel energies. In: Proceedings of the ECCV 2016 Part I, pp. 614–627 (2016)
Lavenant, H.: Harmonic mappings valued in the Wasserstein space. Technical Report. arXiv:1712.07528, arXiv (2017)
Lee, J.M.: Riemannian Manifolds. An Introduction to Curvature. Springer, New York (1997)
Lellmann, J., Lorenz, D.A., Schönlieb, C., Valkonen, T.: Imaging with Kantorovich–Rubinstein discrepancy. SIAM J. Imaging Sci. 7(4), 2833–2859 (2014)
Lellmann, J., Strekalovskiy, E., Koetter, S., Cremers, D.: Total variation regularization for functions with values in a manifold. In: 2013 IEEE International Conference on Computer Vision, pp. 2944–2951 (2013)
McGraw, T., Vemuri, B., Ozarslan, E., Chen, Y., Mareci, T.: Variational denoising of diffusion weighted MRI. Inverse Probl. Imaging 3(4), 625–648 (2009)
Meesters, S., Sanguinetti, G., Garyfallidis, E., Portegies, J., Duits, R.: Fast implementations of contextual PDE’s for HARDI data processing in DIPY. Technical Report, ISMRM 2016 Conference (2016)
Meesters, S., Sanguinetti, G., Garyfallidis, E., Portegies, J., Ossenblok, P., Duits, R.: Cleaning output of tractography via fiber to bundle coherence, a new open source implementation. Technical Report, Human Brain Mapping Conference (2016)
Michailovich, O.V., Rathi, Y.: On approximation of orientation distributions by means of spherical ridgelets. IEEE Trans. Image Process. 19(2), 461–477 (2010)
Miranda, M.: Functions of bounded variation on "good" metric spaces. Journal de Mathématiques Pures et Appliquées 82(8), 975–1004 (2003)
Mollenhoff, T., Laude, E., Moeller, M., Lellmann, J., Cremers, D.: Sublabel-accurate relaxation of nonconvex energies. In: The IEEE Conference on Computer Vision and Pattern Recognition (CVPR) (2016)
MomayyezSiahkal, P., Siddiqi, K.: 3D stochastic completion fields for mapping connectivity in diffusion MRI. IEEE Trans. Pattern Anal. Mach. Intell. 35(4), 983–995 (2013)
Ncube, S., Srivastava, A.: A novel Riemannian metric for analyzing HARDI data. In: Proceedings of the SPIE, p. 7962 (2011)
Ouyang, Y., Chen, Y., Wu, Y.: Vectorial total variation regularisation of orientation distribution functions in diffusion weighted MRI. Int. J. Bioinform. Res. Appl. 10(1), 110–127 (2014)
Pock, T., Chambolle, A.: Diagonal preconditioning for first order primal-dual algorithms in convex optimization. In: 2011 International Conference on Computer Vision, Barcelona, pp. 1762–1769 (2011)
Portegies, J., Duits, R.: New exact and numerical solutions of the (convection–)diffusion kernels on SE(3). Differ. Geom. Appl. 53, 182–219 (2017)
Portegies, J.M., Fick, R.H.J., Sanguinetti, G.R., Meesters, S.P.L., Girard, G., Duits, R.: Improving fiber alignment in HARDI by combining contextual PDE flow with constrained spherical deconvolution. PLOS ONE 10(10), e0138,122 (2015)
Prčkovska, V., Andorrà, M., Villoslada, P., Martinez-Heras, E., Duits, R., Fortin, D., Rodrigues, P., Descoteaux, M.: Contextual diffusion image post-processing aids clinical applications. In: Hotz, I., Schultz, T. (eds.) Visualization and Processing of Higher Order Descriptors for Multi-Valued Data, pp. 353–377. Springer, Berlin (2015)
Reisert, M., Kellner, E., Kiselev, V.: About the geometry of asymmetric fiber orientation distributions. IEEE Trans. Med. Imaging 31(6), 1240–1249 (2012)
Reisert, M., Skibbe, H.: Fiber continuity based spherical deconvolution in spherical harmonic domain. In: Medical Image Computing and Computer-Assisted Intervention—MICCAI 2013, pp. 493–500. Springer, Berlin (2013)
Rokem, A., Yeatman, J., Pestilli, F., Wandell, B.: High angular resolution diffusion MRI. Stanford Digital Repository (2013). http://purl.stanford.edu/yx282xq2090. Accessed 20 Sept 2017
Skibbe, H., Reisert, M.: Spherical tensor algebra: a toolkit for 3d image processing. J. Math. Imaging Vis. 58(3), 349–381 (2017)
Srivastava, A., Jermyn, I.H., Joshi, S.H.: Riemannian analysis of probability density functions with applications in vision. In: CVPR ’07, pp. 1–8 (2007)
Stejskal, E., Tanner, J.: Spin diffusion measurements: spin echos in the presence of a time-dependent field gradient. J. Chem. Phys. 42, 288–292 (1965)
Tax, C.M.W., Jeurissen, B., Vos, S.B., Viergever, M.A., Leemans, A.: Recursive calibration of the fiber response function for spherical deconvolution of diffusion MRI data. NeuroImage 86, 67–80 (2014)
Tournier, J.D., Calamante, F., Connelly, A.: Robust determination of the fibre orientation distribution in diffusion MRI: non-negativity constrained super-resolved spherical deconvolution. NeuroImage 35(4), 1459–1472 (2007)
Tournier, J.D., Calamante, F., Gadian, D., Connelly, A.: Direct estimation of the fibre orientation density function from diffusion-weighted MRI data using spherical deconvolution. NeuroImage 23(3), 1176–1185 (2004)
Tuch, D.S.: Q-ball imaging. Magn. Reson. Med. 52(6), 1358–1372 (2004)
Tuch, D.S., Reese, T.G., Wiegell, M.R., Makris, N., Belliveau, J.W., Wedeen, V.J.: High angular resolution diffusion imaging reveals intravoxel white matter fiber heterogeneity. Magn. Reson. Med. 48(4), 577–582 (2002)
Villani, C.: Optimal Transport. Old and New. Springer, Berlin (2009)
Vogt, T., Lellmann, J.: An optimal transport-based restoration method for Q-ball imaging. In: 2017 Scale Space and Variational Methods in Computer Vision, pp. 271–282 (2017)
Weaver, N.: Lipschitz Algebras. World Scientific, Singapore (1999)
Weinmann, A., Demaret, L., Storath, M.J.: Mumford–Shah and Potts regularization for manifold-valued data. J. Math. Imaging Vis. 55(3), 428–445 (2016)
Author information
Authors and Affiliations
Corresponding author
Appendices
Appendix A: Background from Functional Analysis and Measure Theory
In this appendix, we present the theoretical background for a rigorous understanding of the notation and definitions underlying the notion of \({\text {TV}}\) as proposed in (5) and (7). Section A.1 is concerned with Banach space-valued functions, and Sect. A.2 focuses on the special case of measure-valued functions.
1.1 Banach Space-Valued Functions of Bounded Variation
This subsection introduces a function space on which the formulation of \({\text {TV}}\) as given in (5) is well defined.
Let \((V, \Vert \cdot \Vert _V)\) be a real Banach space with (topological) dual space \(V^*\), i.e., \(V^*\) is the set of bounded linear operators from V to \(\mathbb {R}\). The dual pairing is denoted by \(\langle p, v \rangle := p(v)\) whenever \(p \in V^*\) and \(v \in V\).
We say that \(u:\varOmega \rightarrow V\) is weakly measurable if \(x \mapsto \langle p, u(x) \rangle \) is measurable for each \(p \in V^*\) and say that \(u \in L_w^\infty (\varOmega , V)\) if u is weakly measurable and essentially bounded in V, i.e.,
Note that the essential supremum is well defined even for non-measurable functions as long as the measure is complete. In our case, we assume the Lebesgue measure on \(\varOmega \) which is complete.
The following Lemma ensures that the integrand in (5) is measurable.
Lemma 1
Assume that \(u:\varOmega \rightarrow V\) is weakly measurable and \(p:\varOmega \rightarrow V^*\) is weakly* continuous, i.e., for each \(v \in V\), the map \(x \mapsto \langle p(x), v \rangle \) is continuous. Then, the map \(x \mapsto \langle p(x), u(x) \rangle \) is measurable.
Proof
Define \(f:\varOmega \times \varOmega \rightarrow \mathbb {R}\) via
Then, f is continuous in the first and measurable in the second variable. In the calculus of variations, functions with this property are called Carathéodory functions and have the property that \(x \mapsto f(x,g(x))\) is measurable whenever \(g:\varOmega \rightarrow \varOmega \) is measurable, which is proven by approximation of g as the pointwise limit of simple functions [22, Proposition 3.7]. In our case, we can simply set \(g(x) := x\), which is measurable, and the assertion follows. \(\square \)
1.2 Wasserstein Metrics and the KR Norm
This subsection is concerned with the definition of the space of measures \(K\!R(X)\) and the isometric embedding \(\mathcal {P}(X) \subset K\!R(X)\) underlying the formulation of \({\text {TV}}\) given in (7).
By \(\mathcal {M}(X)\) and \(\mathcal {P}(X) \subset \mathcal {M}(X)\), we denote the sets of signed Radon measures and Borel probability measures supported on X. \(\mathcal {M}(X)\) is a vector space [40, p. 360] and a Banach space if equipped with the norm
so that a function \(u:\varOmega \rightarrow \mathcal {P}(X) \subset \mathcal {M}(X)\) is Banach space-valued (i.e., u takes values in a Banach space). If we define C(X) as the space of continuous functions on X with norm \(\Vert f\Vert _{C} := \sup _{x \in X} |f(x)|\), under the above assumptions on X, \(\mathcal {M}(X)\) can be identified with the (topological) dual space of C(X) with dual pairing
whenever \(\mu \in \mathcal {M}(X)\) and \(p \in C(X)\), as proven in [40, p. 364]. Hence, \(\mathcal {P}(X)\) is a bounded subset of a dual space.
We will now see that additionally, \(\mathcal {P}(X)\) can be regarded as subset of a Banach space which is a predual space (in the sense that its dual space can be identified with a “meaningful” function space) and which metrizes the weak* topology of \(\mathcal {M}(X)\) on \(\mathcal {P}(X)\) by the optimal transport metrics we are interested in.
For \(q \ge 1\), the Wasserstein metrics \(W_q\) on \(\mathcal {P}(X)\) are defined via
where
Here, \(\pi _i\gamma \) denotes the ith marginal of the measure \(\gamma \) on the product space \(X \times X\), i.e., \(\pi _1\gamma (A) := \gamma (A \times X)\) and \(\pi _2\gamma (B) := \gamma (X \times B)\) whenever \(A, B \subset X\).
Now, let \({\text {Lip}}(X,\mathbb {R}^d)\) be the space of Lipschitz-continuous functions on X with values in \(\mathbb {R}^d\) and \({\text {Lip}}(X) := {\text {Lip}}(X,\mathbb {R}^1)\). Furthermore, denote the Lipschitz seminorm by \([\cdot ]_{{\text {Lip}}}\) so that \([f]_{{\text {Lip}}}\) is the Lipschitz constant of f. Note that, if we fix some arbitrary \(x_0 \in X\), the seminorm \([\cdot ]_{{\text {Lip}}}\) is actually a norm on the set
The famous Kantorovich–Rubinstein duality [44] states that, for \(q=1\), the Wasserstein metric is actually induced by a norm, namely \(W_1(\mu , \mu ') = \Vert \mu - \mu '\Vert _{K\!R}\), where
whenever \(\nu \in \mathcal {M}_0(X) := \{ \mu \in \mathcal {M}:\int _X d\mu = 0\}\). The completion \(K\!R(X)\) of \(\mathcal {M}_0(X)\) with respect to \(\Vert \cdot \Vert _{K\!R}\) is a predual space of \(({\text {Lip}}_0(X), [\cdot ]_{{\text {Lip}}})\) [79, Theorem 2.2.2 and Cor. 2.3.5].Footnote 2 Hence, after subtracting a point mass at \(x_0\), the set \(\mathcal {P}(X) - \delta _{x_0}\) is a subset of the Banach space \(K\!R(X)\), the predual of \({\text {Lip}}_0(X)\).
Consequently, the embeddings
define two different topologies on \(\mathcal {P}(X)\). The first embedding space \((\mathcal {M}(X), \Vert \cdot \Vert _{\mathcal {M}})\) is isometrically isomorphic to the dual of C(X). The second embedding space \((K\!R(X), \Vert \cdot \Vert _{K\!R})\) is known to be a metrization of the weak*-topology on the bounded subset \(\mathcal {P}(X)\) of the dual space \(\mathcal {M}(X) = C(X)^*\) [77, Theorem 6.9].
Importantly, while \((\mathcal {P}(X), \Vert \cdot \Vert _{\mathcal {M}})\) is not separable unless X is discrete, \((\mathcal {P}(X), \Vert \cdot \Vert _{K\!R})\) is in fact compact, in particular complete and separable [77, Theorem 6.18] which is crucial in our result on the existence of minimizers (Theorem 1).
Appendix B: Proof of \({\text {TV}}\)-Behavior for Cartoonlike Functions
Proof
(Prop. 1) Let \(p:\varOmega \rightarrow (V^*)^d\) satisfy the constraints in (5) and denote by \(\nu \) the outer unit normal of \(\partial U\). The set \(\varOmega \) is bounded, p and its derivatives are continuous and \(u \in L_w^\infty (\varOmega , V)\) since the range of u is finite and U, \(\varOmega \) are measurable. Therefore, all of the following integrals converge absolutely. Due to linearity of the divergence,
Using this property and applying Gauss’ theorem, we compute
For the last inequality, we used our first assumption on \(\Vert \cdot \Vert _{(V^*)^d}\) together with the norm constraint for p in (5). Taking the supremum over p as in (5), we arrive at
For the reverse inequality, let \({\tilde{p}} \in V^*\) be arbitrary with the property \(\Vert {\tilde{p}}\Vert _{V^*} \le 1\) and \(\phi \in C_c^1(\varOmega , \mathbb {R}^d)\) satisfying \(\Vert \phi (x)\Vert _2 \le 1\). Now, by (11), the function
has the properties required in (5). Hence,
Taking the supremum over all \(\phi \in C_c^1(\varOmega , \mathbb {R}^d)\) satisfying \(\Vert \phi (x)\Vert _2 \le 1\), we obtain
where \({\text {Per}}(U, \varOmega )\) is the perimeter of U in \(\varOmega \). In the theory of Caccioppoli sets (or sets of finite perimeter), the perimeter is known to agree with \(\mathcal {H}^{d-1}(\partial U)\) for sets with \(C^1\) boundary [4, p. 143].
Now, taking the supremum over all \({\tilde{p}} \in V^*\) with \(\Vert {\tilde{p}}\Vert _{V^*} \le 1\) and using the fact that the canonical embedding of a Banach space into its bidual is isometric, i.e.,
we arrive at the desired reverse inequality which concludes the proof. \(\square \)
Appendix C: Proof of Rotational Invariance
Proof
(Proposition 2) Let \(R \in SO(d)\) and define
In (5), the norm constraint on p(x) is equivalent to the norm constraint on \({\tilde{p}}(y)\) by condition (13). Now, consider the integral transform
where, using \(R^T R = I\),
which implies \({\text {TV}}_V(u) = {\text {TV}}_V({\tilde{u}})\). \(\square \)
Appendix D: Discussion of Product Norms
There is one subtlety about formulation (5) of the total variation: The choice of norm for the product space \((V^*)^d\) affects the properties of our total variation seminorm.
1.1 Product Norms as Required in Proposition 1
The following proposition gives some examples for norms that satisfy or fail to satisfy conditions (10) and (11) in Proposition 1 about cartoonlike functions.
Proposition 4
The following norms for \(p \in (V^*)^d\) satisfy (10) and (11) for any normed space V:
-
1.
For \(s = 2\):
$$\begin{aligned} \Vert p\Vert _{(V^*)^d,s} := \left( \sum _{i=1}^d \Vert p_i\Vert _{V^*}^s \right) ^{1/s}. \end{aligned}$$(D.1) -
2.
Writing \(p(v):=(\langle p_1,v\rangle ,\dots ,\langle p_d,v\rangle )\in \mathbb {R}^d\), \(v \in V\),
$$\begin{aligned} \Vert p\Vert _{\mathcal {L}(V, \mathbb {R}^d)} := \sup _{\Vert v\Vert _V \le 1} \Vert p(v)\Vert _{2} \end{aligned}$$(D.2)
On the other hand, for any \(1 \le s < 2\) and \(s > 2\), there is a normed space V such that at least one of the properties (10), (11) is not satisfied by corresponding product norm (D.1).
Remark 1
In the finite-dimensional Euclidean case \(V = \mathbb {R}^n\) with norm \(\Vert \cdot \Vert _2\), we have \((V^*)^d = \mathbb {R}^{d,n}\); thus, p is matrix-valued and \(\Vert \cdot \Vert _{\mathcal {L}(V, \mathbb {R}^d)}\) agrees with the spectral norm \(\Vert \cdot \Vert _\sigma \). The norm defined in (D.1) is the Frobenius norm \(\Vert \cdot \Vert _F\) for \(s=2\).
Proof
(Prop. 4) By Cauchy–Schwarz,
whenever \(p \in (V^*)^d\), \(v \in V\), and \(x \in \mathbb {R}^d\). Similarly, for each \(q \in V^*\),
Hence, for \(s = 2\), properties (10) and (11) are satisfied by product norm (D.1).
For operator norm (D.2), consider
which is property (10). On the other hand, (11) follows from
Now, for \(s > 2\), property (10) fails for \(d = 2\), \(V = V^* = \mathbb {R}\), \(p = x = (1,1)\) and \(v = 1\) since
For \(1 \le s < 2\), consider \(d = 2\), \(V^* = \mathbb {R}\), \(q = 1\) and \(x = (1,1)\), then
which contradicts property (11). \(\square \)
1.2 Rotationally Symmetric Product Norms
For \(V = (\mathbb {R}^n, \Vert \cdot \Vert _2)\), property (13) in Proposition 2 is satisfied by the Frobenius norm as well as the spectral norms on \((V^*)^d = \mathbb {R}^{d,n}\). In general, the following proposition holds:
Proposition 5
For any normed space V, rotational invariance property (13) is satisfied by operator norm (D.2). For any \(s \in [1,\infty )\), there is a normed space V such that property (13) does not hold for product norm (D.1).
Proof
By definition of the operator norm and rotational invariance of the Euclidean norm \(\Vert \cdot \Vert _2\),
For product norms (D.1), without loss of generality, we consider the case \(d = 2\), \(V := (\mathbb {R}^2, \Vert \cdot \Vert _1)\), \(p_1 = (1,0)\), \(p_2 = (0,1)\) and
Then, \(V^* := (\mathbb {R}^2, \Vert \cdot \Vert _\infty )\) and
whereas
for any \(1 \le s < \infty \). \(\square \)
1.3 Product Norms on \({\text {Lip}}_0(X)\)
We conclude our discussion about product norms on \((V^*)^d\) with the special case of \(V = K\!R(X)\): For \(p \in [{\text {Lip}}_0(X)]^d\), the most natural choice is
which is automatically rotationally invariant. On the other hand, the product norm defined in (D.1) (with \(s=2\)), namely \(\sqrt{\sum _{i=1}^d [p_i]_{{\text {Lip}}}^2}\), is not rotationally invariant for general metric spaces X. However, in the special case \(X \subset (\mathbb {R}^n, \Vert \cdot \Vert _2)\) and \(p \in C^1(X,\mathbb {R}^d)\), norms (D.22) and (D.1) coincide with \(\sup _{z\in X} \Vert Dp(z)\Vert _\sigma \) (spectral norm of the Jacobian) and \(\sup _{z\in X} \Vert Dp(z)\Vert _F\) (Frobenius norm of the Jacobian), respectively, both satisfying rotational invariance.
Appendix E: Proof of Non-uniqueness
Proof
(Prop. 3) Let \(u \in L_w^\infty (\varOmega , \mathcal {P}(X))\). With the given choice of X, there exists a measurable function \({\tilde{u}}:\varOmega \rightarrow [0,1]\) such that
The measurability of \({\tilde{u}}\) is equivalent to the weak measurability of u by definition:
The constraint
from the definition of \({\text {TV}}_{K\!R}\) in (7) translates to
Furthermore,
By the compact support of \(p_1\), the last term vanishes when integrated over \(\varOmega \). Consequently,
and therefore
Thus we have shown that the functional \(T_{\rho ,\lambda }\) is equivalent to the classical \(L^1\)-\({\text {TV}}\) functional with the indicator function \(\mathbf {1}_U\) as input data and evaluated at \({\tilde{u}}\) which is known to have non-unique minimizers for a certain choice of \(\lambda \) [17]. \(\square \)
Appendix F: Proof of Existence
1.1 Well-Defined Energy Functional
In order for the functional defined in (15) to be well defined, the mapping \(x \mapsto \rho (x, u(x))\) needs to be measurable. In the following lemma, we show that this is the case under mild conditions on \(\rho \).
Lemma 2
Let \(\rho :\varOmega \times \mathcal {P}(X) \rightarrow [0,\infty )\) be a globally bounded function that is measurable in the first and convex in the second variable, i.e., \(x \mapsto \rho (x,\mu )\) is measurable for each \(\mu \in \mathcal {P}(X)\), and \(\mu \mapsto \rho (x,\mu )\) is convex for each \(x \in \varOmega \). Then, the map \(x \rightarrow \rho (x,u(x))\) is measurable for every \(u \in L_w^\infty (\varOmega , \mathcal {P}(X))\).
Remark 2
As will become clear from the proof, the convexity condition can be replaced by the assumption that \(\rho \) be continuous with respect to \((\mathcal {P}(X), W_1)\) in the second variable. However, in order to ensure weak* lower semicontinuity of functional (15), we will require convexity of \(\rho \) in the existence proof (Theorem 1) anyway. Therefore, for simplicity we also stick to the (stronger) convexity condition in Lemma 2.
Remark 3
One example of a function satisfying the assumptions in Lemma 2 is given by
Indeed, boundedness follows from the boundedness of the Wasserstein metric in the case of an underlying bounded metric spaces (here \(\mathbb {S}^2\)). Convexity in the second argument follows from the fact that the Wasserstein metric is induced by a norm (A.8).
Proof
(Lemma 2) The metric space \((\mathcal {P}(X), W_1)\) is compact, hence separable. By Pettis’ measurability theorem [10, Chapter VI, §1, No. 5, Proposition 12], weak and strong measurability coincide for separably valued functions, so that u is actually strongly measurable as a function with values in \((\mathcal {P}(X),W_1)\). Note, however, that this does not imply strong measurability with respect to the norm topology of \((\mathcal {M}(X), \Vert \cdot \Vert _{\mathcal {M}})\) in general!
As bounded convex functions are locally Lipschitz continuous [19, Theorem 2.34], \(\rho \) is continuous in the second variable with respect to \(W_1\). As in the proof of Lemma 1, we now note that \(\rho \) is a Carathéodory function, for which compositions with measurable functions such as \(x \mapsto \rho (x,u(x))\) are known to be measurable. \(\square \)
1.2 The Notion of Weakly* Measurable Functions
Before we can go on with the proof of existence of minimizers to (15), we introduce the notion of weak* measurability because this will play a crucial role in the proof.
Analogously with the notion of weak measurability and with \(L_{w}^\infty (\varOmega , K\!R(X))\) introduced above, we say that a measure-valued function \(u:\varOmega \rightarrow \mathcal {M}(X)\) is weakly* measurable if the mapping
is measurable for each \(f \in C(X)\). \(L_{w*}^\infty (\varOmega , \mathcal {M}(X))\) is defined accordingly as the space of weakly* measurable functions.
For functions \(u:\varOmega \rightarrow \mathcal {P}(X)\) mapping onto the space of probability measures, there is an immediate connection between weak* measurability and weak measurability: u is weakly measurable if the mapping
is measurable whenever \(p \in {\text {Lip}}_0(X)\). However, since, by the Stone–Weierstrass theorem, the Lipschitz functions \({\text {Lip}}(X)\) are dense in \((C(X), \Vert \cdot \Vert _{\infty })\) [13, p. 198], both notions of measurability coincide for probability measure-valued functions \(u:\varOmega \rightarrow \mathcal {P}(X)\), so that
However, as this equivalence does not hold for the larger spaces \(L_{w*}^\infty (\varOmega , \mathcal {M}(X))\) and \(L_{w}^\infty (\varOmega , \mathcal {M}(X))\), it will be crucial to keep track of the difference between weak and weak* measurability in the existence proof.
1.3 Proof of Existence
Proof
(Theorem 1) The proof is guided by the direct method from the calculus of variations. The first part is inspired by the proof of the fundamental theorem for Young measures as formulated and proven in [6].
Let \(u^k:\varOmega \rightarrow \mathcal {P}(X)\), \(k \in \mathbb {N}\), be a minimizing sequence for \(T_{\rho ,\lambda }\), i.e.,
As \(\mathcal {M}(X)\) is the dual space of C(X), \(L_{w*}^\infty (\varOmega , \mathcal {M}(X))\) with the norm defined in (A.1) is dual to the Banach space \(L^1(\varOmega , C(X))\) of Bochner integrable functions on \(\varOmega \) with values in C(X) [42, p. 93]. Now, \(\mathcal {P}(X)\) as a subset of \(\mathcal {M}(X)\) is bounded so that our sequence \(u^k\) is bounded in \(L_{w*}^\infty (\varOmega , \mathcal {M}(X))\) (here we use again that \(L_{w*}^\infty (\varOmega , \mathcal {P}(X)) = L_{w}^\infty (\varOmega , \mathcal {P}(X))\)).
Note that we get boundedness of our minimizing sequence “for free”, without any assumptions on the coercivity of \(T_{\rho ,\lambda }\)! Hence we can apply the Banach–Alaoglu theorem, which states that there exist \(u^\infty \in L_{w*}^\infty (\varOmega , \mathcal {M}(X))\) and a subsequence, also denoted by \(u^k\), such that
Using the notation in (A.4), this means by definition
We now show that \(u^\infty (x) \in \mathcal {P}(X)\) almost everywhere, i.e., \(u^\infty \) is a nonnegative measure of unit mass: Convergence (F.7) holds in particular for the choice \(p(x,s) := \phi (x)f(s)\), where \(\phi \in L^1(\varOmega )\) and \(f \in C(X)\). For nonnegative functions \(\phi \) and f, we have
for all k, which implies
Since this holds for all nonnegative \(\phi \) and f, we deduce that \(u^\infty (x)\) is a nonnegative measure for almost every \(x \in \varOmega \). The choice \(f(s) \equiv 1\) in (F.7) shows that \(u^\infty \) has unit mass almost everywhere.
Therefore, \(u^\infty (x) \in \mathcal {P}(X)\) almost everywhere and we have shown that \(u^\infty \) lies in the feasible set \(L_{w}^\infty (\varOmega , \mathcal {P}(X))\). It remains to show that \(u^\infty \) is in fact a minimizer.
In order to do so, we prove weak* lower semicontinuity of \(T_{\rho ,\lambda }\). We consider the two integral terms in definition (15) of \(T_{\rho ,\lambda }\) separately. For the \({\text {TV}}_{K\!R}\) term, for any \(p \in C_c^1(\varOmega , {\text {Lip}}(X,\mathbb {R}^d))\), we have \({\text {div}}p \in L^1(\varOmega , C(X))\) so that
Taking the supremum over all p with \([p(x)]_{[{\text {Lip}}(X)]^d} \le 1\) almost everywhere, we deduce lower semicontinuity of the regularizer:
The data fidelity term \(u \mapsto \int _\varOmega \rho (x,u(x)) \,\hbox {d}x\) is convex and bounded on the closed convex subset \(L_w^\infty (\varOmega , \mathcal {P}(X))\) of the space \(L_{w*}^\infty (\varOmega , \mathcal {M}(X))\). It is also continuous, as convex and bounded functions on normed spaces are locally Lipschitz continuous. This implies weak* lower semicontinuity on \(L_w^\infty (\varOmega , \mathcal {P}(X))\).
Therefore, the objective function \(T_{\rho ,\lambda }\) is weakly* lower semicontinuous, and we obtain
for the minimizing sequence \((u^k)\), which concludes the proof.
\(\square \)
Rights and permissions
About this article
Cite this article
Vogt, T., Lellmann, J. Measure-Valued Variational Models with Applications to Diffusion-Weighted Imaging. J Math Imaging Vis 60, 1482–1502 (2018). https://doi.org/10.1007/s10851-018-0827-8
Received:
Accepted:
Published:
Issue Date:
DOI: https://doi.org/10.1007/s10851-018-0827-8